PRPDPaD

bmIdwtt

aotonwfa

i

OmhAi[rqosf

drd

t

A

ADBc

©P

AD

ULT

CA

RD

IAC

reoperative Cardiac Troponin I to Assess Midtermisks of Coronary Bypass Grafting Operations inatients With Recent Myocardial Infarctionomenico Paparella, MD, Giuseppe Scrascia, MD, Andreas Paramythiotis, MD,ietro Guida, PhD, Vito Magari, MD, Pietro Giorgio Malvindi, MD, Stefano Favale, MD,nd Luigi de Luca Tupputi Schinosa, MD

ivisions of Cardiac Surgery and Cardiology, Department of Emergency and Organ Transplant, University of Bari, Bari, Italy

oa(pattair

icim

Background. The optimal timing for coronary arteryypass grafting (CABG) in patients with recent acuteyocardial infarction (AMI) is unclear. Cardiac troponin(cTnI) is a widely accepted biomarker of myocardialamage. The objective of this study was to determinehether preoperative cTnI values could be used to de-

ermine risk stratification for CABG operations in pa-ients with recent AMI.

Methods. Evaluated were 184 patients who sustainedn AMI within 21 days of undergoing nonurgent CABGperations. They were divided into two groups accordingo their preoperative cTnI values: 117 patients with cTnIf 0.15 ng/mL or less and 67 with cTnI exceeding 0.15g/mL. Associations between study variables and eventsere assessed with logistic regression modelling. Time

rom AMI to operation was evaluated to define preoper-tive cTnI variation.Results. Values of cTnI tended to decrease when the

nterval between AMI and the operation increased. Pre-

Cm(dOouaoctc

imibTtecpw

ari, Piazza Giulio Cesare 11, 70100 Bari, Italy; e-mail: dpaparella@ardiochir.uniba.it.

2010 by The Society of Thoracic Surgeonsublished by Elsevier Inc

perative cTnI values were significantly associated withhigher incidence of major postoperative complications

low cardiac output syndrome, intraaortic balloonump necessity, mechanical ventilation >72 hours,cute renal failure, in-hospital mortality). Periopera-ive myocardial damage was more pronounced in pa-ients with cTnI exceeding 0.15 ng/mL. Multivariatenalyses revealed cTnI exceeding 0.15 ng/mL was anndependent predictor for 6-month mortality (oddsatio, 3.7; p � 0.043).

Conclusions. Preoperative cTnI exceeding 0.15 ng/mLn patients with recent AMI undergoing CABG is asso-iated with higher postoperative myocardial damage ands a strong determinant of postoperative morbidity and

ortality within the 6-month period.

(Ann Thorac Surg 2010;89:696–703)

© 2010 by The Society of Thoracic Surgeons

ptimal timing of coronary artery bypass grafting(CABG) operations in patients with recent acute

yocardial infarction (AMI) is unclear. Several studiesave demonstrated that CABG performed close to anMI event carries a higher risk of postoperative morbid-

ty and death than operations performed at a later date1, 2]. However, the reason why delaying the CABGenders the operation safer has not been explained. Theuestion remains: What is the optimal time for CABGperations in patients with recent AMI under relativelytable clinical conditions and an angiographic indicationor surgical revascularization?

Cardiac troponins have become the gold standard foretecting myocardial damage [3] and for diagnosis andisk stratification in patients with acute coronary syn-rome (ACS) [4, 5].Cardiac troponin assessments have also been adopted

o define the perioperative risks in patients undergoing

ccepted for publication Nov 30, 2009.

ddress correspondence to Dr Paparella, Division of Cardiac Surgery,ipartimento d’Emergenza e Trapianti d’Organo (D.E.T.O.), University of

ABG after recent ACS. A few studies have docu-ented that elevated preoperative cardiac troponin I

cTnI) levels are associated with increased operativeeath and adverse events after emergency CABG [6 –9].ther studies have demonstrated that increased post-

perative myocardial damage, evaluated by cTnI val-es, is associated with increased postoperative deathnd morbidity [10 –12]. The influence of elevated pre-perative cTnI values on midterm postoperative out-ome has not been reported nor has the reason behindhe negative prognostic value of preoperative elevatedTnI been clearly elucidated.

The primary aim of the present study was to verify thenfluence of preoperative cTnI values on the incidence of

ajor postoperative complications and 6-month mortal-ty of patients who sustained an AMI within 21 daysefore undergoing CABG in stable clinical conditions.he relationship between preoperative cTnI values and

he interval between AMI and the operation has beenvaluated. We also documented the effect of preoperativeTnI values on postoperative cTnI release to identify theotential pathophysiology of adverse outcome in patients

ith a recent AMI.

0003-4975/10/$36.00doi:10.1016/j.athoracsur.2009.11.072

M

Tbow

tsJAfipcprp

PTrcwua

rwip

tqwnfdrfp

dcw0w0t[ep

T

V

PMABHDHSSCUECCPRA

E

P

BE

697Ann Thorac Surg PAPARELLA ET AL2010;89:696–703 PREOPERATIVE CARDIAC TROPININ I IN CABG

AD

ULT

CA

RD

IAC

aterial and Methods

he approval of the local ethic committee was obtainedefore our study was conducted, and in accordance withur local ethic committee policy, individual consentsere waived.This was a single-center, retrospective, and observa-

ional study that used clinical and laboratory data pro-pectively recorded in our institutional database. Fromanuary 2005 to December 2007, 184 patients with recentMI (onset � 21 days) underwent elective, isolated,rst-time CABG operations with cardiopulmonary by-ass. The study excluded emergency operations (eg,ardiogenic shock, intraaortic balloon pump [IABP], oratients otherwise requiring an immediate operation),epeat operations, combined procedures, and off-pumprocedures.

atientshe patients arrived at Policlinico Hospital of Bari di-ectly in the Emergency Department or from differentardiology units without coronary angiography facilitiesithin the geographic area. AMI was suspected by eval-ating symptoms and electrocardiogram abnormalitiesnd confirmed by cTnI measurements. Coronary angiog-

able 1. Preoperative Patient Characteristics Based on Preope

ariable All Patients

atients, totals 184ales, No. (%) 139 (76)ge, mean � SD, y 65 � 9MI, mean � SD, kg/m2 27 � 4ypertension, % 69iabetes Mellitus, % 43ypercholesterolemia, % 71tatin therapy, % 54moking history, % 54OPD, % 21nstable angina, % 35jection fraction, mean � SD 0.45 � 0.09A stenosis � 50%, % 10erebrovascular accident, % 4eripheral vascular disease, % 11enal disease, % 5cute myocardial infarction, No. (%)�7 days 46 (25)8–14 days 43 (23)�14 days 95 (52)

uroSCORE, mean � SDLogistic 8.94 � 9.65Additive 6.65 � 2.93

re-op cTnI, mean � SD, ng/mL 1.19 � 3.61

MI � body mass index; CA � carotid artery; COPD � chronic obstructuropean System for Cardiac Operative Risk Evaluation; SD � standard d

aphy was performed in the Division of Cardiologyithin 48 hours of the onset of AMI. The operations were

ndicated by the severity of coronary stenosis or byostinfarction angina, or both.Patients were considered nonemergency cases because

hey had stable clinical conditions and were subse-uently transferred to the Division of Cardiac Surgeryithout discharge home. The waiting period from diag-osis to operation depended on a variety of casual

actors, including the surgical waiting list, concomitantisease needing preoperative evaluation, and preangiog-aphy aggressive antiplatelet therapy. The exact timerom AMI to operation was evaluated by reviewingatients’ medical records.For the purposes of our study, these patients were

ivided in two groups according to their preoperativeTnI value. The low cTnI group consisted of 117 patientsho had preoperative cTnI of 0.15 ng/mL or less (cTnI �

.15 ng/mL). The high cTnI group consisted of 67 patientsho had preoperative cTnI exceeding 0.15 ng/mL (cTnI �

.15 ng/mL). A cTnI value of 0.15 ng/mL is the value usedo confirm the diagnosis of AMI at our Pathology Unit13]. Clinical management was not influenced by preop-rative cTnI values neither perioperatively nor after hos-ital discharge.

e Cardiac Troponin I Values

Pre-op cTnI, ng/mL

�0.15 �0.15 p Value

117 6793 (79) 46 (69) 0.1064 � 9 66 � 9 0.2426 � 4 27 � 4 0.10

67 73 0.3644 43 0.9773 67 0.4357 48 0.2158 46 0.1225 15 0.1223 55 �0.001

0.45 � 0.09 0.44 � 0.09 0.809 10 0.824 4 0.95

11 12 0.866 4 0.66

�0.00112 (10) 34 (51)26 (22) 17 (25)79 (68) 16 (24)

7.27 � 7.08 11.85 � 12.52 0.0026.05 � 2.63 7.70 � 3.15 �0.0010.03 � 0.04 3.21 � 5.44 . . .

rativ

ive pulmonary disease; cTnI � cardiac troponin I; EuroSCORE �eviation.

STemioslcfto

DH1PdtdPntc

odiwssbiu

sbogeditewg

SOHwuaTMottt

vw

TSlwmdNpp

FApt

TO

V

I

P

a

cir

698 PAPARELLA ET AL Ann Thorac SurgPREOPERATIVE CARDIAC TROPININ I IN CABG 2010;89:696–703A

DU

LTC

AR

DIA

C

tudy Objectiveshe primary objective of our study was to evaluate theffect of preoperative cTnI values on the incidence ofajor postoperative complications and 6-month mortal-

ty rates in clinically stable patients undergoing CABGperations. A major postoperative complication is con-idered the occurrence of at least one of the following:ow cardiac output syndrome, IABP necessity, mechani-al ventilation lasting longer than 72 hours, acute renalailure, and in-hospital death. The second objective waso verify the effect of preoperative cTnI values on post-perative cTnI values.

efinitionsypertension was defined as blood pressure exceeding

40/90 mm Hg or needing antihypertensive medications.atients who had a history of diabetes were considerediabetic. Hypercholesterolemia was defined as choles-

erol level greater than 200 mg/dL. Smoking history wasefined as any current or past form of tobacco use.reoperative renal disease was defined as serum creati-ine exceeding 2.0 mg/dL. On-going refractory angina

hat required the use of intravenous nitrate therapy forontrol was regarded as unstable angina.

Death within the same hospital admission, regardlessf the cause, was defined as operative mortality. Cardiaceath was regarded as any death due to cardiac causes,

ncluding sudden death. Low cardiac output syndromeas defined as the need for postoperative inotropic

upport or an IABP for more than 12 hours to maintainystolic blood pressure greater than 90 mm Hg, meanlood pressure greater than 60 mm Hg, or the cardiac

ndex greater than 2.2 L/min/m2, despite sufficient vol-me substitution.Extubation criteria were hemodynamic stability, ab-

ence of surgical bleeding, consciousness, and optimallood gas measurements with a fraction of inspiredxygen of 0.3 or less. Cerebrovascular disease was re-arded as any transient ischemic attack, reversible isch-mic neurologic deficit, or stroke. Acute renal failure wasefined as new onset of postoperative creatinine exceed-

ng 2.0 mg/dL or an increase of creatinine levels greaterhan twofold compared with preoperative creatinine lev-ls or requirement of dialysis. Allogenic red blood cellsere transfused if the hemoglobin value was less than 8/dL.

urgical Managementur surgical approach was always a median sternotomy.eparin was given (300 U/Kg), and CPB was establishedith ascending aorta and two-stage venous cannulationsing moderate hypothermia (34°C), a centrifugal pump,nd uncoated tubing system with membrane oxygenator.ranexamic acid (2 g) was added to prime solution.yocardial protection was achieved using antegrade and

ptional retrograde cold blood cardioplegia. Intraopera-ive heparin monitoring was by standard activated clot-ing time (ACT, Medtronic Inc, Minneapolis, MN). Addi-

ional heparin boluses (5000 U) were given if the ACT

ci

alues were less than 400 seconds. Protamine sulphateas administered to reverse heparin.

roponin Measurementamples for preoperative cTnI measurements were col-

ected the morning of the operation. Cardiac troponin Ias measured according to the manufacturer’s recom-endation by standard immunoassay technique (Car-

iac Troponin-I Flex Reagent Cartridge, Dade Behring,ewark, NJ). Postoperative cTnI measurements wereerformed at 6, 12, 24, and 36 hours and then daily untilostoperative day 6.

ollow-Upfter discharge home, follow-up was accomplished byeriodic evaluation of patients at our institution or by

elephone contact with the patients’ general practitio-

able 2. Intraoperative Variables and Postoperativeutcomes Based on Preoperative Cardiac Troponin I Values

Pre-op cTnI, ng/mL

ariablea �0.15 �0.15 p Value

ntraoperativeGrafts, No. 2.95 � 1.06 2.85 � 1.06 0.64Arterial grafts, No. 1.01 � 0.32 1.06 � 0.37 0.65Operation time, h 4.76 � 1.03 4.93 � 1.1 0.60IAPB use 2 (2) 8 (12) 0.005CPB time, min 95 � 50 96 � 47 0.95Cross-clamp time, min 50 � 27 49 � 25 0.96Blood products

transfusion, U0.33 � 0.84 0.13 � 0.52 0.57

ostoperative outcomesIntensive care unit

stay, h39 � 13 43 � 19 0.07

Overall stay, d 9.8 � 9.2 11.8 � 10.9 0.19Mechanical ventilation

� 4 h8 (7) 11 (16) 0.04

Inotropes use 15 (13) 10 (14) 0.69IABP useb 2 (2) 8 (12) 0.003Total blood loss, mL 771 � 261 808 � 309 0.52Blood products

transfusion, U1.64 � 3.25 2.00 � 2.76 0.57

Acute renal failure 11 (9) 4 (6) 0.58LCOS 15 (13) 15 (22) 0.09Atrial fibrillation 27 (23) 25 (37) 0.039Sepsis 2 (2) 1 (1) 1.00cTnI peak, ng/mL 7.01 � 9.63 24.49 � 48.02 �0.001In-hospital mortality 5 (4) 6 (9) 0.20Major post-op

complicationsc21 (18) 21 (31) 0.037

6-month mortality 5 (4) 9 (13) 0.024

Continuous variables are expressed as the mean � standard deviation;ategoric variables as number (%). b IABP insertion after CPB wean-ng. c LCOS, IABP necessity, mechanical ventilation � 72 hours, acuteenal failure, in-hospital mortality.

TnI � cardiac troponin I; CPB � cardiopulmonary bypass; IABP �ntraaortic balloon pump; LCOS � low cardiac output syndrome.

nc

SDToppFav

wor(cacaMlsp

R

TmawTbdap

bteIIocclw

Ftc

T

F

MABHDHSCUECPRcc

BT

699Ann Thorac Surg PAPARELLA ET AL2010;89:696–703 PREOPERATIVE CARDIAC TROPININ I IN CABG

AD

ULT

CA

RD

IAC

ers. Follow-up was stopped at 6 months and was 100%omplete.

tatistical Analysisata are shown as mean values � standard deviation.he continuous variables were compared using analysisf variance or t test for independent samples. The non-arametric Mann-Whitney U test was used when appro-riate. Categoric data were tested with the �2 test or theisher exact test, where appropriate. cTnI was evaluateds continuous and as a dichotomous variable (�0.15s �0.15 ng/mL). Multivariate logistic regression model

ig 1. Kaplan-Meier curves are shown for patients with a preopera-ive cardiac troponin I (cTnI) � 0.15 ng/mL (dashed line) andTnI � 0.15 ng/mL (solid line).

able 3. Logistic Regression Analysis for Major Postoperative

actor

Yes No

(n � 42) (n � 142)

ales, No. (%) 29 (69) 110 (77)ge, mean � SD, y 68 � 8 64 � 9MI, mean � SD, kg/m2 26 � 4 27 � 4ypertension, % 71 68iabetes mellitus, % 45 43ypercholesterolemia, % 76 69moking history, % 48 56OPD, % 21 21nstable angina, % 40 33F, mean � SD 0.41 � 0.11 0.46 � 0.08A stenosis � 50%, % 10 10VD, % 17 10enal disease, % 17 2TnI, mean � SD, ng/mL 2.53 � 5.40 0.79 � 2.77TnI � 0.15 ng/mL, % 50 32

MI � body mass index; CA � carotid artery; CI � confidence interval;roponin I; EF � ejection fraction; OR � odds ratio; PVD � periph

as applied including variables with a significance levelf less than 0.20 at univariate analysis. Estimated oddsatios (OR) and corresponding 95% confidence intervalsCI) and probability values are reported. The OR for aontinuous variable refers to the risk ratio per unit of thenalyzed variable unless specified otherwise. Survivalurves at 6 months between the cTnI � 0.15 ng/mL groupnd the cTnI � 0.15 ng/mL group were based on Kaplan-eier analyses and were compared by means of the

og-rank test. The analyses were made using Statistica 6.1oftware (StatSoft Inc, Tulsa, OK), and values of� 0.05 were considered statistically significant.

esults

able 1 reports the preoperative characteristics and de-ographics of the patients in our study group. Preoper-

tive use of continuous intravenous nitrate and heparinas significantly higher in the cTnI � 0.15 ng/mL group.his influenced the preoperative risk stratification byoth the logistic and additive European System for Car-iac Operative Risk Evaluation (EuroSCORE). Intraoper-tive variables and postoperative outcomes according toreoperative cTnI values are reported in Table 2.The CABG operations were performed similarly in

oth groups, and no material differences in intraopera-ive variables were noted. The cTnI � 0.15 ng/mL groupxperienced longer mechanical ventilation time, moreABP use, and a higher incidence of atrial fibrillation.ntensive care unit length of stay, overall hospital lengthf stay, and low cardiac output syndrome were higher inTnI � 0.15 ng/mL patients, without statistically signifi-ant differences. Mechanical ventilation was significantlyonger in cTnI�0.15 ng/mL patients. In-hospital mortalityas 4% in the cTnI � 0.15 ng/mL group and 9% in

plications

Univariate Multivariate

R (95% CI) p Value OR (95% CI) p Value

5 (0.30–1.40) 0.27 . . . . . .5 (1.01–1.10) 0.016 1.03 (0.98–1.09) 0.1820 (0.81–1.00) 0.044 0.90 (0.80–1.01) 0.0776 (0.54–2.49) 0.70 . . . . . .1 (0.55–2.20) 0.79 . . . . . .4 (0.65–3.20) 0.37 . . . . . .2 (0.36–1.45) 0.36 . . . . . .2 (0.43–2.38) 0.97 . . . . . .7 (0.67–2.80) 0.38 . . . . . .4 (0.90–0.98) 0.002 0.95 (0.90–0.99) 0.0146 (0.30–3.12) 0.95 . . . . . .3 (0.68–4.91) 0.23 . . . . . .7 (2.26–38.02) 0.002 8.34 (1.55–44.90) 0.0132 (1.02–1.23) 0.019 1.20 (1.04–1.39) 0.0141 (1.03–4.22) 0.039 . . . . . .

Com

O

0.61.00.91.1

1.1.40.71.01.30.90.91.89.21.12.8

COPD � chronic obstructive pulmonary disease; cTnI � cardiaceral vascular disease; SD � standard deviation.

citc0Kt0f

Wcp1nma3n

osfith9

TFmwwblannwa1

C

Agismaaam

Ffcc

T

F

MABHDHSCUECPRcc

B erval;t riphe

700 PAPARELLA ET AL Ann Thorac SurgPREOPERATIVE CARDIAC TROPININ I IN CABG 2010;89:696–703A

DU

LTC

AR

DIA

C

TnI � 0.15 patients, which was not significant. Thencidence of major postoperative complications andhe 6-month mortality rate were significantly higher inTnI � 0.15 ng/mL patients (31% and 13%) than in cTnI �.15 ng/mL patients (18% and 4%). Figure 1 showsaplan-Meier curves for the 6-month mortality rate of

he groups with preoperative TnI � 0.15 ng/mL vs cTnI �.15 ng/mL. All deaths that occurred during the 6-monthollow-up were due to cardiac causes.

Multivariable analyses are reported in Tables 3 and 4.hen used as a continuous variable, the preoperative

TnI value was an independent predictor for majorostoperative complications (OR, 1.20; 95% CI, 1.04 to.39; p � 0.014 per unit of cTnI increase). However, it didot predict 6-month mortality. When used as a dichoto-ous variable, abnormal preoperative cTnI values were

ssociated with increased risk of death at 6 months (OR,.74; 95% CI, 1.04 to 13.48; p � 0.043 per cTnI � 0.15g/mL).

ig 2. Mean values with the standard error (error bars) are shownor cardiac troponin I (cTnI) in patients with preoperative valuesTni � 0.15 ng/mL (open circle) and cTnI � 0.15 ng/mL (closed

able 4. Logistic Regression Analysis for 6-Month Mortality

actor

Yes No

(n � 14) (n � 170)

ales, No. (%) 8 (57) 131 (77)ge, mean � SD, y 69 � 8 65 � 9MI, mean � SD, kg/m2 26 � 5 27 � 4ypertension, % 79 68iabetes mellitus, % 64 42ypercholesterolemia, % 64 71moking history, % 36 55OPD, % 21 21nstable angina, % 43 34F, mean � SD 0.36 � 0.11 0.45 � 0.09A stenosis � 50%, % 14 9VD, % 21 11enal disease, % 7 5TnI, mean � SD, ng/mL 2.60 � 6.16 1.07 � 3.31TnI � 0.15 ng/mL, % 64 34

MI � body mass index; CA � carotid artery; CI � confidence introponin I; EF � ejection fraction; OR � odds ratio; PVD � pe

iircle) through postoperative day 6.

As expected, postoperative cTnI values increased aftern-pump CABG in all patients; however, they wereignificantly higher in the cTnI � 0.15 ng/mL group in therst 36 postoperative hours and in postoperative days 2

o 5 (Fig 2). Postoperative cTnI mean peak values wereigher in cTnI � 0.15 group patients (24.5 � 48.0 vs 7.01 �.63 ng/mL; p � 0.0002).

ime From AMIigure 3 shows cTnI mean values and the 6-monthortality rate in patients who underwent operationsithin 1, 2, or 3 weeks after AMI. Mean cTnI valuesere lower in patients with a longer delay from AMIefore operation. The 6-month mortality rate was

ower in patients operated on during the second weekfter AMI but the difference was not statistically sig-ificant. Considering the cTnI � 0.15 ng/mL patients,o significant difference in the 6-month mortality rateas observed between patients who underwent oper-

tions before or after 7 days from AMI (respectively,5% and 12%; p � 0.52).

omment

n increasing number of patients receive coronary an-iography in the context of recent AMI. This has led to an

ncrease in candidates for CABG operations as a re-ponse to failed angioplasty or because of left main/ultivessel disease. Most of these patients do not require

n immediate operation but are commonly transferred asmatter of urgency to cardiac surgical wards. Surgeons

re therefore faced with the difficult decision of deter-ining the optimal timing of CABG operations for clin-

Univariate Multivariate

R (95% CI) p Value OR (95% CI) p Value

0 (0.13–1.22) 0.11 0.39 (0.08–1.88) 0.247 (1.00–1.14) 0.06 1.09 (1.00–1.19) 0.066 (0.83–1.11) 0.58 . . . . . .1 (0.45–6.43) 0.43 . . . . . .1 (0.80–7.87) 0.11 2.30 (0.62–8.50) 0.213 (0.23–2.30) 0.59 . . . . . .5 (0.14–1.41) 0.17 0.63 (0.12–3.32) 0.582 (0.27–3.81) 0.98 . . . . . .5 (0.48–4.40) 0.51 . . . . . .0 (0.85–0.96) �0.001 0.87 (0.81–0.94) �0.0010 (0.33–7.89) 0.56 . . . . . .0 (0.58–9.12) 0.23 . . . . . .8 (0.16–11.89) 0.77 . . . . . .7 (0.97–1.19) 0.16 . . . . . .8 (1.11–10.93) 0.032 3.74 (1.04–13.48) 0.043

COPD � chronic obstructive pulmonary disease; cTnI � cardiacral vascular disease; SD � standard deviation.

O

0.41.00.91.72.50.70.41.01.40.91.62.31.31.03.4

cally stable patients with a recent AMI.

ahssaatmittrAtC“Agf

tbup

pInso

ecTwwan

aidiowmmhannctvavt

cfwtttMcptdonstacti0

pdebo

Fspi

701Ann Thorac Surg PAPARELLA ET AL2010;89:696–703 PREOPERATIVE CARDIAC TROPININ I IN CABG

AD

ULT

CA

RD

IAC

Even with improvements in myocardial protection andnesthetic management, patients with recent AMI stillave a high mortality rate after CABG operations. Retro-pective examination shows that CABG operationshould be postponed, when possible, for 3 or more daysfter the onset of the AMI. A retrospective multicenternalysis of 44,365 patients who underwent CABG after aransmural or nontransmural AMI revealed that hospital

ortality decreased in all patients with increasing wait-ng time from AMI. Patients with transmural and non-ransmural AMI had similar postoperative outcomes, buthe mortality rate was higher in patients with a transmu-al AMI if CABG was performed within 7 days after theMI [1]. Weiss and colleagues [2] evaluated 40,159 pa-

ients hospitalized for AMI who underwent subsequentABG. Patients were stratified by the timing of CABG inearly” (�2 days from AMI) and “late” (�3 days fromMI) groups. Mortality rates were higher in the early-roup patients, suggesting that CABG ought to be deferredor 3 or more days after AMI in nonurgent patients [2].

Preoperative cardiac troponin levels are more accuratehan a “time from AMI” evaluation and therefore haveeen introduced to improve risk stratification of patientsndergoing CABG. Thielmann and colleagues [14] used

ig 3. (A) Mean values with the standard error (error bars) arehown for cardiac troponin I (cTnI) and (B) 6-month mortality inatients operated on within week 1, 2, or 3 after acute myocardialnfarction. (ANOVA � analysis of variance).

reoperative cTnI values to divide 1405 elective CABG s

atients in three groups: group I, cTnI � 0.1 ng/ml; groupI, cTnI 0.11 to 1.15 ng/mL; and group III, cTnI � 1.5g/mL. The group III patient experienced longer inten-ive care unit stay and higher incidences of low cardiacutput syndrome and in-hospital mortality.The prognostic value of preoperative cTnI was also

valuated also in 57 patients with ST-elevation AMI andompared with 197 patients with non-ST-elevation AMI.he study concluded that ST elevation patients had aorse clinical outcome and that preoperative cTnI valueas an independent determinant of in-hospital mortality

nd major adverse coronary events both in ST andon-ST-elevation patients [8].We focused our attention in this study on patients withrecent MI who did not require emergency revascular-

zation because their clinical condition was stable. Weemonstrated that preoperative cTnI values were an

ndependent predictor for the incidence of major post-perative complications and cTnI exceeding 0.15 ng/mLas associated with an increased risk of death during 6onths of follow-up. We showed that it takes generallyore than 14 days for cTnI to return to normal values;

owever, cTnI is still elevated in some patients 2 weeksfter AMI. This is probably due to the extension of theecrosis that occurred during MI; unfortunately, we wereot able to analyze peak post-AMI cTnI values or loss ofontractility after MI. However, we found it interestinghat the mortality risk was similar in patients with ele-ated cTnI levels who underwent operations before orfter 7 days from AMI, suggesting that preoperative cTnIalues may be a more precise prognostic indicator thanhe interval between AMI and the operation.

Preoperative and intraoperative patient variables, in-luding strong outcome predictors such as age, ejectionraction, cardiopulmonary bypass, and cross-clamp time,ere similar in both groups and therefore suggested that

he reason why postoperative prognosis is poor in pa-ients with elevated preoperative cTnI values may be dueo a higher degree of perioperative myocardial damage.

yocardial damage, evaluated by serial postoperativeTnI measurements, is in fact significantly increased inatients with preoperative cTnI values that are higher

han normal levels. Increased postoperative myocardialamage influences not only the immediate postoperativeutcome but also 6-month mortality rates. Using mag-etic resonance imaging, Steuer and colleagues [15]howed that elevated cTnI, creatine kinase-MB, androponin T levels after on-pump CABG correspond to themount of perioperatively infarcted myocardium. Theorrelation between the release of cardiac enzymes andhe loss of vital myocardium could probably explain thencreased mortality rate that we observed in the cTnI �.15 ng/mL group.Different surgical strategies or pharmacologic com-

ounds, or both, may reduce perioperative myocardialamage in the general population undergoing CABG; forxample, antegrade/retrograde cardioplegia [16], warmlood cardioplegia [17], preoperative IABP [18], and pre-perative levosimendan administration [19]. Future re-

earch should be focused on identifying the optimal

stw

ttelao

oiscwtuava

Wm

R

1

1

1

1

1

1

1

1

1

1

I

Atiafdudmwr

702 PAPARELLA ET AL Ann Thorac SurgPREOPERATIVE CARDIAC TROPININ I IN CABG 2010;89:696–703

©P

AD

ULT

CA

RD

IAC

urgical or pharmacologic approach to limit periopera-ive myocardial damage in patients undergoing CABGith ongoing laboratory signs of recent AMI.The principle limitation of our study was its retrospec-

ive design and the consequent lack of serial postopera-ive echocardiographic or magnetic resonance imagingvaluation of left ventricular function to possibly corre-ate postoperative myocardial function with survival. Aslready stated, another important limitation was the lackf a cTnI value when AMI was initially diagnosed.We therefore conclude that, if possible, normalization

f cTnI values before CABG operations seems warrantedn patients who have sustained a recent AMI, but thishould be confirmed by specifically designed studies. Iflinical conditions or coronary anatomy require a patientith elevated cTnI values to undergo a CABG operation,

hen surgeons ought to be aware that these patients arender a higher risk of perioperative myocardial damagend, consequently, a higher risk of postoperative ad-erse events, including death, within the 6 monthsfter the operation.

e thank Felicia Kohn Passaro for kindly reviewing theanuscript.

eferences

1. Lee DC, Oz MC, Weinberg AD, Lin SX, Ting W. Optimaltiming of revascularization: transmural versus nontransmu-ral acute myocardial infarction. Ann Thorac Surg 2001;71:1198–204.

2. Weiss ES, Chang DD, Joyce DL, Nwakanma LU, Yuh DD.Optimal timing of coronary artery bypass after acute myo-cardial infarction: a review of California discharge data.J Thorac Cardiovasc Surg 2008;135:503–11.

3. Adams JE, Bodor GS, Davila Roman VG, et al. Cardiactroponin I: a marker with a high specificity for cardiac injury.Circulation 1993;88:101–6.

4. Alpert JS, Thygesen K, Antman E, Bassand JP. Myocardialinfarction redefined—a consensus document of The JointEuropean Society of Cardiology/American College of Cardi-ology Committee for the redefinition of myocardial infarc-tion. J Am Coll Cardiol 2000;36:959–69.

5. Antman EM, Tanasijevic MJ, Thompson B, et al. Cardiac-specific troponin I levels to predict the risk of mortality inpatients with acute coronary syndromes. N Engl J Med

1996;335:1342–9.

isky, and too late may invite another acute coronary

elo

itsnmpr

2010 by The Society of Thoracic Surgeonsublished by Elsevier Inc

6. Thielmann M, Neuhauser M, Marr A, et al. Predictors andoutcomes of coronary artery bypass grafting in st elevationmyocardial infarction. Ann Thorac Surg 2007;84:17–24.

7. Thielmann M, Massoudy P, Neuhauser M, et al. Prognosticvalue of preoperative cardiac troponin I in patients withnon-ST-segment elevation acute coronary syndromes un-dergoing coronary artery bypass surgery. Chest 2005;128:3526–36.

8. Thielmann M, Massoudy P, Neuhauser M, et al. Prognosticvalue of preoperative cardiac troponin I in patients under-going emergency coronary artery bypass surgery with non-ST-elevation or ST-elevation acute coronary syndromes.Circulation 2006;114(suppl I):I-448–53.

9. Carrier M, Pelletier LC, Martineau R, Pellerin M, SolymossBC. In elective coronary artery bypass grafting, preoperativetroponin T level predicts the risk of myocardial infarction.J Thorac Cardiovasc Surg 1998;115:1328–34.

0. Paparella D, Cappabianca G, Visicchio G, et al. Cardiactroponin I release after coronary artery bypass graftingoperation: effects on operative and midterm survival. AnnThorac Surg 2005;80:1758–64.

1. Paparella D, Cappabianca G, Malvindi P, et al. Myocardialinjury after off-pump coronary artery bypass grafting oper-ation. Eur J Cardiothorac Surg 2007;32:481–7.

2. Lasocki S, Provenchère S, Bénessiano J, et al. Cardiac tropo-nin I is an independent predictor of in-hospital death afteradult cardiac surgery. Anesthesiology 2002;97:405–11.

3. Amodio G, Antonelli G, Varraso L, Ruggieri V, Di Serio F.Clinical impact of the troponin 99th percentile cut-off andclinical utility of myoglobin measurement in the early man-agement of chest pain patients admitted to the EmergencyCardiology Department. Coron Artery Dis 2007;18:181–6.

4. Thielmann M, Massoudy P, Neuhauser M, et al. Risk strat-ification with cardiac troponin I in patients undergoingelective coronary artery bypass surgery. Eur J CardiothoracSurg 2005;27:861–9.

5. Steuer J, Bjerner T, Duvernoy O, et al. Visualisation andquantification of peri-operative myocardial infarction aftercoronary artery bypass surgery with contrast-enhancedmagnetic resonance imaging. Eur Heart J 2004;25:1293–9.

6. Onorati F, De Feo M, Mastroroberto P, et al. Determinantsand prognosis of myocardial damage after coronary arterybypass grafting. Ann Thorac Surg 2005;79:837–45.

7. Mazzetti A, Calafiore AM, Lapenna D, et al. Intermittentantegrade warm cardioplegia reduces oxidative stress andimproves metabolism of the ischemic-reperfused humanmyocardium. J Thorac Cardiovasc Surg 1995;109:787–95.

8. Dyub AM, Whitlock RP, Abouzhr LL, Cinà CS. Preoperativeintra-aortic balloon pump in patients undergoing coronarybypass surgery: a systematic review and meta-analysis.J Card Surg 2008;23:79–86.

9. Tritapepe L, De Santis V, Vitale D, et al. Precondotioningeffects of levosimendan in coronary artery bypass grating, a

pilot study. Br J Anaesth 2006;96:694–700.

NVITED COMMENTARY

s cardiac surgeons, we are quite frequently faced withhe situation of performing coronary artery bypass graft-ng after acute myocardial infarction (AMI). With thedvent of primary angioplasty, the patients referred to usor surgical intervention will be those who have beeneemed unfit for percutaneous intervention because ofnfavorable anatomy of coronary disease for angioplasty,iffuse disease, or poor left ventricular function. Theost important clinical decision in these situations ishen to perform the operation. Too soon may be too

vent. Various studies, as quoted by Paparella and col-eagues [1], have tried to inform us about the risk of theperation according to the time elapsed since the AMI.It is appreciated that authors in this study have tried to

nvestigate whether one can have a guideline based onhe measurement of cardiac troponin I (cTnI). In thistudy, the patients with a cTnI level of less than 0.15g/mL had an advantage in terms of less postoperativeorbidity and death at 6 months. Assuming that now we

ut this in practice, would we wait to perform surgical

evascularization in patients with AMI until the cTnI falls

0003-4975/10/$36.00doi:10.1016/j.athoracsur.2009.12.055